Tuesday 13 April 2021

Synthesizer Build part-42: 8 RANDOM GATES by Yusynth.

 Creates 8 random gate outputs from one gate input signal which can be as high in frequency as an audio signal. Lots of creative possibilities with this module.

There is an other random gates project on my website already. That one is included in the Noise Module article and it creates random pulses on one output. With this module we have 8 different outputs which trigger in a completely random order. It needs a squarewave on the input that can come from an LFO, the gate out from a sequencer, the clock pulse from a sample and hold or even the pulse wave output from a VCO. To quote the YuSynth website: "If feeding the GATE IN with a high frequency pulse coming from a VCO, each GATE output becomes an individual coloured digital noise source usable for sound effects. The colour of the noise will directly depend on the frequency from the VCO. White noise is obtained for frequencies above 30kHz".. 

ABOUT THE CIRCUIT:
The module is fed with only positive voltage so no dual powersource needed. It works fine on both +15V or +12V. You can feed the gate-in with signals that have a negative cycle to them. It will simply ground the negative part of the cycle through diode D1. The output gate signals have an amplitude of 8 Volt when powered from a +15V powersupply.
This build consists mainly of wirebridges. My layout has 37 of them. All the output stages are made on separate pieces of stripboard with just 4 strips of 10 holes. They are soldered straight to the output sockets. I did this to save space otherwise I would have had to make a separate print with all the outputs on them. This way saves space and also hookup wire. The three 100nF capacitors you can see on the layout are meant to be de-coupling caps but where they are positioned is really too far away from the chips to be effective. So instead of putting them where the layout shows them, solder them straight over the plus and ground pins of the IC's (top right and bottom left of each chip).

Here's the layout I made. First the wiring diagram:


(Last revised: 14-April-2021: Added missing 1K resistor to output prints.)

In the box, on the wiring diagram above, you can see the schematic drawing of the output stripboards. I left out the 1K resistors in series with the output in my original design. I had simply forgotten it but I have now updated everything and the 1K resistor is now included. It helps to protect the transistor against short circuits, smooths the output voltage a bit and also determins the output impedance.
The 270 Ohm and the LED together with the 1K resistor to ground form a voltage divider that determins the voltage of the outputed Gate signal. That voltage is normally 8 Volt but if you want it to be higher you can make the 1K to ground a higher value like 1K5 or lower for a lower output Gate voltage.

Here's the main stripboard. It's only 24 by 48 holes but you could try to redesign it and make it even more compact so it would fit in a Eurorack system. For instance, if you connected the outputs straight to the correct pins of the chip instead of using the wirebridges you can save about 8 or 9 holes in width. Certainly enough room to make it fit a Eurorack system. And because it's a "Random" gates generator, the correct order doesn't really matter does it?


And here's a close-up of the little output stripboard that is soldered straight to the output socket: (If you print this one, choose the A6 format to save some printer ink.)
You need to make 8 of these output prints. Some cuts are a bit hard to see but the top two strips are cut at position 5 and there's an other cut at position C8.


(Last revised: 14-April-2021: Added missing 1K resistor to output prints.)

Here's the Bill of Materials:



Here's the schematic by Yusynth. You can find the original YuSynth article by clicking HERE.


As you can see it's actually quite a simple circuit. It mostly consists of connections between the three IC's. It is mentioned on the YuSynth website that this module needs a bit of time before it starts behaving correctly. When you first start it up it will probably not fire on all cilinders and display a repeating pattern with only about 4 or 5 LEDs lighting up and after at least ten cycles this will change into a random pattern using all the outputs. However, since I changed the new CD4070 I had in there for a used vintage CD4070 from the 1980's that I had lying around, the module works good right from the start. 
The module that I built was at first prone to hanging. It would suddenly stop being random and get stuck in a 4 or 5 LED pattern. Only by changing the input gate frequency or pulling the Gate-In cable in and out a few times would I get it working again. It turned out this was also due to the IC's I was using. I don't know if it was a fake chip or if it was damaged but I changed IC-3 for an old stock CD4070 that I once de-soldered out of an organ circuit board and the problem was solved immediately.
So make sure the chips you're using come from a reputable source!

MODE SWITCH:
There's a mode switch that lets you choose between two settings. In the ON position the output stays high until it detects the next pulse, so the pulses don't have any dead time between them. In the OFF position the output pulse stops on the negative slope of the input gate pulse, so the output pulses will have the same length as the input gate pulses.
There's also an option to advance the pulses manually with a momentary switch (normally off). This switch is connected to the internal switch of the Gate input socket so it will only work when there is no cable connected to the gate input socket.

Some screenshots from the oscilloscope. The first one shows how extremely fast the risetime of the output gate signals is. Just over 123 nanno seconds! That's 0 to 8 Volt in 0.000000123 seconds. This means theoretically that it could handle signals upto 40MHz! (Agreed, this knowledge is of no use in the synthesizer world but it fascinates me personally because I also have a background in radio technology and transmitters ^___^).


Here's what the output sequence of one of the random gates looks like. A non-repeating sequence of pulses with an amplitude of 8V. 



Here are some pictures of the build proces:

Wirebridges. In this picture there's a little wirebridge missing connecting pins 7 and 8 of IC2 (CD4051).


Here's the finished print. Like I mentioned earlier, the de-coupling caps are much to far away from the chips to be effective so get some small ceramic 100nF caps and carefully solder them straight over the plus and minus connections of the chips on the copper side. I myself left it like this and it works just fine because I don't use a switchmode powersupply but a linear one, with a big transformer. 


Here's the main board with the 8 output prints. My output prints are missing the 1K resistor in series with the output sockets (I had forgotten those) but they are included in the layouts. That 1K resistor helps to make the output waves smoother. I could see that on the oscilloscope images. It also protects the transistors by limiting the current going through them should the output be shorted. (Although damage will be very unlikely even without the 1K resistors because the pulses are so short).


Finished panel backside wiring:


Frontal view of the mounted panel:


And here's a little test video showing the module firing randomly on all cilinders :)


TIP:
If you want a fast pulse train with random gaps in it, then connect 4 outputs from this module to the 4 inputs of a mixer, like the mixer/passive attenuator module on this website. At the output of the mixer you will get a pulse train with random gaps in them. It's cool to use this on the cut-off of a filter to add some random spice to the sound.
If you then set the switch on the random gates module to 'Stay high until the next gate pulse' you have sort of a random voltage generator, although there will still be random 0V gaps in the output but that makes it unique :)
You could even make a little TL072 mixer print and include it in this module. Choose how many inputs you want (less than 8 of course) and connect those mixer-inputs to whichever outputs you choose and then make an extra output socket on the panel that carries the output from the mixer and label it "Pulse Train". It's just a thought but there are many ways to adapt this design to your own needs.

Okay, that's an other one done. If you have any questions or remarks please put them in the comments below or post on the special Facebook Group for this website where we have a great community of synth enthousiasts willing to help you.

If you successfully built this module and you're using it in a cool way that others might enjoy, please make a video, put it on YouTube and contact me with the link. I'll add it to the article with full credit given.

4 comments:

  1. Hey Eddy,

    Thanks for the write ups, lots of good info in here for a first time synth builder like myself. I think I may have caught on error though:

    I think you have the On/Off functions descriptions backwards. (And it looks like the YuSynth article confirms this). "On" should have the gate outputs gated by the current state of switch Q1 by passing the voltage over Q1 to the inhibit pin of the 4051. Meanwhile "Off" pulls inhibit to ground, which should keep the gate outputs latched at whatever their last state was on the previous clock cycle. (Inhibit is active high on the 4051).

    Just wanted to double check here, maybe I'm the one who has it mixed up!

    ReplyDelete
    Replies
    1. Oh, that's well possible that I mixed that up. It wouldn't be the first time. Anyway, once you turn this on you can easily determine which is which and label it accordingly :)

      Delete
  2. Hello Eddy.
    I am new to the DIY module area and i am currently facing a problem with this module. Problem child is the 10 ohm resistor just after the power supply. This drives immediately, after switching on, my room temperature up. However, the smoke development bothers me :P Maybe you can help me there. Do I have to change components when I increase the resistance?

    ReplyDelete
    Replies
    1. Just leave that 10 Ohm out and see if everything works. It functions as a noise rejector (sort off). It is weird that it goes up in smoke because there shouldn't be that much current going through it. Take it out and put in a wire bridge and take it from there.

      Delete

Note: comments are moderated and do not appear straightaway. Your first comment is not allowed to contain any links.